Construction and Comprehensive Analysis of ceRNA Networks and Tumor-Infiltrating Immune Cells in Hepatocellular Carcinoma With Vascular Invasion

Shijiao Cai,Renle Du,Yuan Zhang,Zhengyi Yuan,Jie Shang,Yang Yang,Bin Han,Weilong Zhong,Hengjie Yuan,Zhengxiang Li

DOI: https://doi.org/10.3389/fbinf.2022.836981

2022-01-01

Frontiers in Bioinformatics

Abstract:Background: Hepatocellular carcinoma (HCC) is a common malignant cancer. Metastasis plays a critical role in tumor progression, and vascular invasion is considered one of the most crucial factors for HCC metastasis. However, comprehensive analysis focusing on competitive endogenous RNA (ceRNA) and immune infiltration in the vascular invasion of HCC is lacking. Methods: The gene expression profiles of 321 samples, including 210 primary HCC cases and 111 HCC cases with vascular invasion, were downloaded from The Cancer Genome Atlas-Liver Hepatocellular Carcinoma project, and used in identifying significant differentially expressed lncRNAs (DElncRNAs), miRNAs (DEmiRNAs), and mRNAs (DEmRNAs). The RNAs associated with vascular invasion were used in constructing a ceRNA network. A multigene-based risk signature was constructed using the least absolute shrinkage and selection operator algorithm. We detected the fractions of 28 immune cell types in HCC through single-sample gene set enrichment analysis (ssGSEA). Finally, the relationship between the ceRNA network and immune cells was determined through correlation analysis and used in clarifying the potential mechanism involved in vascular invasion. Results: Overall, 413 DElncRNAs, 27 DEmiRNAs, and 397 DEmRNAs were recognized in HCC. A specific ceRNA network based on the interaction among 3 lncRNA–miRNA pairs and 24 miRNA–mRNA pairs were established. A ceRNA-based prognostic signature was constructed and used in dividing samples into high- and low-risk subgroups. The signature showed significant efficacy; its 3- and 5-year areas under the receiver operating characteristic curves were 0.712 and 0.653, respectively. ceRNA and ssGSEA integration analysis demonstrated that PART1 ( p = 0, R = −0.33) and CDK5R2 ( p = 0.01, R = −0.15) were negatively correlated to natural killer cells. Conclusion: This study demonstrated that vascular invasion in HCC might be related to PART1, and its role in regulating CDK5R2 and NK cells. A nomogram was developed to predict the prognosis of patients with HCC and demonstrated the value of the ceRNA network and tumor-infiltrating immune cells value in improving personalized management.